Window Grouping

1. A method of automatically determining groups of windows, where the windows are managed by a window managing system, where the window managing system maintains representations of respective windows, and where each representation includes a title of its window for displaying in a title bar of the window, the method comprising: analyzing the titles of the windows to determine semantic associations between the titles; and associating the windows based on the associations between their titles; wherein the analyzing and associating are performed by at least one processor.

2. A method according to claim 1 , wherein the analyzing the titles comprises determining clusters of keywords in the titles and associating the windows based on the clusters.

3. A method according to claim 1 , wherein the analyzing the titles comprises preprocessing the titles, extracting features from the preprocessed titles, and clustering the features in accordance with a probabilistic latent semantic indexing (PLSI) model, and determining the associations between the titles based on the clustered features of the titles.

4. A method according to claim 3 , wherein the preprocessing comprises filtering application-specific words from the titles.

5. A method according to claim 1 , further comprising: determining temporal relations between the windows; and further associating the windows based on the determined temporal relations between the windows.

6. A method according to claim 5 , wherein the temporal relations comprise display relations between windows and switches of focus between windows.

7. A method according to claim 6 , wherein the display relations comprise co-visibility of windows.

8. The method of claim 1 , further comprising: collecting window events associated with the windows, the collected window events including switches of focus between windows; modeling states of the windows based on the collected window events; and assigning various ones of the windows to different groups based on the modeled states and semantic associations between the window titles.

9. The method of claim 8 , wherein the collected window events indicate changes in how the windows are displayed on a display, and wherein the modeled states track changes in states of the windows over a period of time.

10. A computer readable storage medium having instruction stored thereon that, when executed by at least one processor, cause the processor to perform: automatically determining groups of windows, where the windows are managed by a window managing system, where the window managing system maintains representations of respective windows, and where each representation includes a title of its window for displaying in a title bar of the window; the automatically determining groups of windows including analyzing the titles of the windows to determine semantic associations between the titles, and associating the windows based on the associations between their titles.

11. The medium of claim 10 , wherein the analyzing the titles of the windows includes determining clusters of keywords in the titles and associating the windows based on the clusters.

12. The medium of claim 10 , wherein the analyzing the titles comprises: preprocessing the titles, the preprocessing including filtering application-specific words from the titles; extracting features from the preprocessed titles; clustering the features in accordance with a probabilistic latent semantic indexing (PLSI) model; and determining the semantic associations between the titles based on the clustered features of the titles.

13. The medium of claim 10 , further comprising instructions stored thereon that, when executed by at least one processor, cause the processor to perform: determining temporal relations between the windows; and further associating the windows based on the determined temporal relations.

14. The medium of claim 10 , wherein the temporal relations comprise display relations between windows and switches of focus between windows, and wherein the display relations comprise co-visibility of windows.

15. The medium of claim 10 , further comprising instructions stored thereon that, when executed by at least one processor, cause the processor to perform: collecting window events associated with the windows, the collected window events including switches of focus between windows; modeling states of the windows based on the collected window events; and assigning various ones of the windows to different groups based on the modeled states and semantic associations between the window titles.

16. The medium of claim 15 , wherein the collected window events indicate changes in how the windows are displayed on a display, and wherein the modeled states track changes in states of the windows over a period of time including frequencies of switches between windows.

17. A system comprising: at least one processor; and a memory operatively connected to the at least one processor and comprising stored therein instructions that, when executed by the at least one processor, cause the at least one processor to automatically determine groups of windows by analyzing the titles of the windows to determine semantic associations between the titles, and by associating the windows based on the associations between their titles; wherein the windows are managed by a window managing system, the window managing system maintains representations of respective windows, and each representation includes a title of its window for displaying in a title bar of the window.

18. The system of claim 17 , wherein the analyzing the titles of the windows includes determining clusters of keywords in the titles and associating the windows based on the clusters.

19. The system of claim 17 , wherein the memory comprises instructions stored therein that, when executed by the at least one processor, cause the at least one processor to: determine temporal relations between the windows, the temporal relations including switches of focus between windows; and further associate the windows based on the determined temporal relations.

20. The system of claim 17 , wherein the memory comprises instructions stored therein that, when executed by the at least one processor, cause the at least one processor to: collect window events associated with the windows, the collected window events including switches of focus between windows; model states of the windows based on the collected window events; and assign various ones of the windows to different groups based on the modeled states and semantic associations between the window titles.

21. The system of claim 20 , wherein the collected window events include window events that indicate changes in how the windows are displayed on a display, and window events passed between application programs and a window management system that manages windows for arbitrary application programs.